perf(hf-space): pre-load model at module init (Option 3 refactor)

fix(hf-space): lower ZEROGPU_DURATION_SECONDS 120 → 60

fix(parser): bump quoted_span limit 200 → 400 chars

fix(hf-space): drop trust_remote_code=True (Phi-4-mini native phi3 path)

fix: address parallel code-review findings (security + quality + spec)

fix(security): pass user Anthropic key directly to SDK, never via env

test(hf-space): cover _call_huggingface (full) + _call_zerogpu (refactor)

chore(compounding-test): hide HuggingFace API option from dropdowns

fix(hf-space): use provider="auto" for InferenceClient (new routing layer)

fix(hf-space): actionable error for HF inference model_not_supported

fix(hf-space): cap ZeroGPU duration at 120s + actionable HF_TOKEN error

fix(hf-space): trust_remote_code for Phi-4-mini + verbose error reporting

fix(samples): pad all 4 sample initiatives above the 200-word minimum

feat(hf-space): user-friendly copy + API key field + 4 sample initiatives

feat(hf-space): add ZeroGPU backend for HuggingFace Pro Spaces

feat(hf-space): add HuggingFace backend (Gemma/Phi-4) alongside Anthropic

feat(hf-space): Gradio diagnostic with Anthropic SDK + prompt caching (US4)